Level 3: Lacking direct evidence
The Spine Patient Outcomes Research Trial (SPORT) enrolled 607 patients with degenerative spondylolisthesis in either a randomized group (if consenting to randomization) or an observational group (if consenting to follow-up but selecting a treatment course without randomization) (N Engl J Med. 2007;356:2257-2270). Participants were at least 18 years of age and had neurogenic claudication or radicular leg pain with neurologic signs, spinal stenosis demonstrated on cross-sectional imaging, and degenerative spondylolisthesis for at least 12 weeks.
The surgery consisted of standard posterior decompressive laminectomy with or without bilateral single-level fusion (iliac-crest bone grafting with or without posterior pedicle screw instrumentation). Nonsurgical treatment included physical therapy, nonsteroidal anti-inflammatory drugs if tolerated, and education/counseling with at-home exercise instruction.
Of the 304 patients in the randomization group, only 57%, or 159, of those randomized to surgery had surgery at one year (64% at two years), whereas 44% of the 145 patients randomized to nonsurgical treatment had surgery at one year (49% at two years). Of 303 patients in the observational group, 173 chose surgery (97% had surgery at one and two years) and 130 patients chose nonsurgical treatment, but 17% had surgery at one year and 25% at two years.
In an analysis of the randomized trial, there were no significant differences in primary outcomes. Due to high crossover rates, the authors combined the randomized and nonrandomized cohorts and analyzed the results as an observational study according to treatment actually received: 372 patients had surgery and 235 patients had nonsurgical treatment. Comparing surgery vs. nonsurgical treatment at two years, surgery was associated with greater mean changes in bodily pain score (29.9 vs. 11.7, P .05, 0-100 scale, with higher score indicating less severe symptoms), and disability index (-24.2 vs. -7.5, P <.05, 0-100 scale, with lower score indicating less severe symptoms). More patients in the surgery group were very or somewhat satisfied with symptoms (68.8% vs. 32.2%, NNT 3) and 74.1% vs. 24.1% self-rated major improvement (NNT 2).
Overall, this report shows that patients who chose surgery had better outcomes, but it does not demonstrate that the surgery itself (rather than patient selection) was responsible.